View Tweet
1/ Short thread on what we're likely to see *if* James Wynn's monster $1bn position is liquidated.
2/ Caveat. I'm going off public docs, there's always the possibility that there's functionality for these extreme scenarios that aren't in the docs, or like the JellyJelly scenario, the validators hard fork to add new functionality to handle an extreme scenario.
3/ Since BTC is a 40X market, liquidation will occur when the remaining margin is 1.25% of the position margin. At $1.1bn, this would represent about $14mn of equity on the position.
4/ What happens first in liquidation, is the system will attempt to clear as much as possible using the resting liquidity in the book. (There is technically 20% partial liquidation first, but the size and leverage is so large, we can proably assume instant cascade into full liq)
5/ The liquidator system will market sell as much size as it can, without sending the position into negative equity. Since the equity at liquidation will be 1.25%, the amount liquidated this way is basically bound by how much size the book can absorb at 1.25% slippage
6/ Of course the state of the book can change between now and then, but based on current liquidity amounts it looks like the BTC order book can absorb maybe $100-200mn of the position. Which leaves about $800mn-$900mn of notional left to absorb.
7/ The next step in the waterfall is for the HLP vault to absorb the remaining. This would mean HLP takes a $800mn long position on BTC.
(Caveat, it's not entirely clear if the vault *has* to backstop the full amount or if there are max limits in place that lead to ADL)
(Caveat, it's not entirely clear if the vault *has* to backstop the full amount or if there are max limits in place that lead to ADL)
8/ Right now, the entire HLP vault has $335mn of capital. Since BTC is a 40X margin, it could comfortably absorb this position in a working capital sense. However the liquidator sub-strategy only has $7.5mn, so it would have insufficient margin to take the full $800mn position.
9/ Let's assume for now HLP auto-rotates more capital into the liquidator vault and absorbs the whole position. (If not the system would auto fail over into ADL, more on that below).
With $335mn of margin this would be a very large and volatile 2X long BTC position.
With $335mn of margin this would be a very large and volatile 2X long BTC position.
10/ However from a margin/liquidation perspective HLP would still be pretty safe at least in the short term. With $335m of margin, its own liquidation price would be somewhere around $70k.
We can assume that it will start aggressively trying to exit risk well before.
We can assume that it will start aggressively trying to exit risk well before.
11/ Some people are worried about a liquidation cascade , which blows through the entire HL book and temporarily causes the price to wick down to this level or lower
However even if that happens, it alone won't trigger liquidations because of HL's robust price indices
However even if that happens, it alone won't trigger liquidations because of HL's robust price indices
12/ Liquidations are based off the mark price. And the mark price uses a conservative median approach that's designed to protect against HL specific liquidity events.
In a HL specific liquidation cascade, we can assume #2 (the price in the book basically) is the extreme value.
In a HL specific liquidation cascade, we can assume #2 (the price in the book basically) is the extreme value.
13/ If we assume no flash crash contagion to #3 Binance, OKX, Bybit, etc. (which seems like fair assumption, given the liquidation cascade is specific to HL), then the robust price indicator is using #1, the oracle price + EMA of the basis.
14/ This EMA will eventually adjust and trigger liquidations, but the point is it's slow to adjust. Not only would we have to see a liquidation cascade, but we'd have to see the book stay depressed for minutes on end with no arbitrageurs stepping into buy. Probably unlikely
15/ However even if the vault isn't liquidated, as it exits the position it will put continuous pressure on the order book for a while. Therefore we're likely to see the HL price of BTC trade at a persistent discount to Binance et al. for a good while.
16/ The way funding is computed is by taking the difference between the HL price and the oracle price (i.e. Binance et al.) If the HL price remains depressed from HLP sell pressure, then funding should become substantially positive.
17/ One tailwind for the long position in the HLP vault (and any other longs) is they'll be continuously earning large amounts of funding. The flip side is shorts will be bleeding funding at high rates, which increase their pressure to sell.
18/ That in turn increases bid liquidity, which accelerates the rate that the vault can exit its position. Both things are good because it can exit faster, reducing risk and will have positive PnL tailwinds.
19/ That being said, there's still a very large naked long on a volatile asset, so even with these tailwinds, there's meaningful risk of significant drawdown.
The other factor to consider is whether this leads to contagion outside HL, which causes sells at other exchanges.
The other factor to consider is whether this leads to contagion outside HL, which causes sells at other exchanges.
20/ That leaves the final option of ADL (auto deleveraging). If the vault reaches its risk limits or ends up being liquidated, then the remainder of the position will be force closed by picking large and levered shorts on the other side and force closing their positions.
Generated by Thread Navigator
Press ⌘ + S to quick-export